This content is not included in your SAE MOBILUS subscription, or you are not logged in.

Artist-Centric New HMI Software Development Workflow: Development of Real-Time 3D Rendering Engine for Reconfigurable Instrument Clusters

Journal Article
2013-01-0425
ISSN: 1946-4614, e-ISSN: 1946-4622
Published April 08, 2013 by SAE International in United States
Artist-Centric New HMI Software Development Workflow: Development of Real-Time 3D Rendering Engine for Reconfigurable Instrument Clusters
Sector:
Citation: Nakata, K., Kajita, K., Nishikawa, R., Matsumoto, S. et al., "Artist-Centric New HMI Software Development Workflow: Development of Real-Time 3D Rendering Engine for Reconfigurable Instrument Clusters," SAE Int. J. Passeng. Cars – Electron. Electr. Syst. 6(2):397-410, 2013, https://doi.org/10.4271/2013-01-0425.
Language: English

Abstract:

Instrument clusters that display all information on a TFT-LCD screen, also known as reconfigurable instrument clusters, have become the new trend in automotive interiors. DENSO mass-produced the world's first reconfigurable instrument cluster in 2008. To satisfy customer requirements, large quantities of resources were required. Coupled with an iterative process due to requirement changes, development costs became very high. Reducing development costs was vital in order to expand the reconfigurable instrument cluster products line. One solution was to use existing human machine interface (HMI) tools. However, most HMI tools are geared toward software developers and not graphic artists. Furthermore, each tool has its own unique method for image and scene creation, creating an ineffective and sometimes difficult environment for artists familiar with industry-leading computer graphics (CG) software to learn and use the tools. As a result, the artist-to-target workflow suffered from inefficiencies because software developers had to remodel the artists designs created with the CG software. Each time the design was changed, both artists and developers had to reiterate the development process.
The instrument cluster is not only a primary source of information to the driver, but also connects the driver to the rest of the vehicle, establishing a somewhat personal relationship between human and machine. It is imperative that the instrument cluster design reflect clarity, precision and beauty. An artist-centric approach was proposed to reduce the development effort by introducing a data converter and real-time 3D rendering engine. The goal was to achieve minimal loss in the development cycle, from concept to production. The artist's idea, as expressed with industry-leading CG software (Autodesk® 3ds Max®), must be precisely replicated at the target level. The new proposed workflow was as followed: First, an artist created a 3D model and scene in 3D CG software. Next, the 3D model and scene were exported and converted for use in the rendering engine. The converted data included parameters and shader code for the rendering engine. Finally, the data was run on the rendering engine. The rendering engine called the appropriate OpenGL ES commands based on the converted data from the artist and dynamic input data from the developer's application code. With the new workflow, any changes the artist makes are immediately and seamlessly applied to the target.
It is important that the rendering engine interprets the converted data on the embedded platform as close as possible to the artist's intended design. Using automatic shader code generation and frame rate improvement techniques contributed to rendering a high quality image on a TFT-LCD screen. The data converter was able to generate and optimize the shader code automatically based on parameters set by the artist in the 3D CG software.
The new workflow was evaluated during a mass production development project and resulted in improved software development efficiency. While DENSO has succeeded in automating graphics development with data conversion and a rendering engine, the next goal is to integrate a model-based development tool to improve development efficiency.